Method for image registration utilizing particle swarm optimization

1. A system for image registration utilizing particle swarm optimization, the system comprising one or more processors that are configured to perform operations of: selecting a set of image windows from a test image; transforming each image window from the test image, such that a transformation of each image window aligns each image window with a reference image having a center, resulting in a set of transformed image windows; configuring a plurality of software agents to operate as a cooperative swarm to optimize an objective function, wherein each agent is assigned an initial velocity vector to explore a multi-dimensional solution space, where each agent is configured to perform at least one iteration, the iteration being a search in the multi-dimensional solution space for a potential objective function where each agent keeps track of a first position vector representing a current individual best solution that the agent has identified, and a second position vector used to store the current global best solution among all agents; evaluating an objective function at the location of each agent, wherein the objective function represents a measure of registration quality between the set of transformed image windows and the reference image; and comparing the current global best solution found by all of the agents with an optimum solution, wherein if the global best solution is within a predetermined threshold of the optimum solution, then the global best solution represents the registration; and wherein the objective function is defined as: J = ( 1 N ⁢ ∑ k = 1 N ⁢ λ k ⁡ ( I max - d _ k ) , where N denotes the number of image windows, Σ represents a summation, λ k denotes the fraction of overlap of a kth image window with the reference image, I max is the maximum pixel value of the test image, and d k denotes the average absolute difference between the image window and the reference image.

3. A system for image registration utilizing particle swarm optimization as set forth in claim 2 , wherein the system is further configured to perform operations of applying a Gaussian filter to the test image and the reference image to assist the convergence of the plurality of software agents.

4. A system for image registration utilizing particle swarm optimization as set forth in claim 3 , wherein the system is further configured to perform operations of applying a translation to each image window in the set of image windows, the set of image windows comprising a center-of-gravity, such that the center-of-gravity of the set of image windows coincides with the center of the reference image.

5. A system for image registration utilizing particle swarm optimization as set forth in claim 4 , wherein the system is further configured to perform operations of generating an image pyramid of both the test image and the reference image, each image pyramid having a plurality of levels comprising images, wherein each level of each image pyramid is an identical image having a different size and resolution.

6. A system for image registration utilizing particle swarm optimization as set forth in claim 5 , wherein the plurality of software agents are configured to explore each level of each image pyramid in search of the objective function, wherein the agents begin at a top level of each image pyramid and continue down each image pyramid until convergence is reached at a lowest level of each image pyramid.

7. A system for image registration utilizing particle swarm optimization as set forth in claim 6 , wherein the evaluation of the objective function is carried out at the same image pyramid levels using the reference image and a set of image windows extracted from the test image pyramid.

8. A computer-implemented method for image registration utilizing particle swarm optimization, the method comprising an act of causing a processor to perform operations of: selecting a set of image windows from a test image; transforming each image window from the test image, such that a transformation of each image window aligns each image window with a reference image having a center, resulting in a set of transformed image windows; configuring a plurality of software agents to operate as a cooperative swarm to optimize an objective function, wherein each agent is assigned an initial velocity vector to explore a multi-dimensional solution space, where each agent is configured to perform at least one iteration, the iteration being a search in the multi-dimensional solution space for a potential objective function where each agent keeps track of a first position vector representing a current individual best solution that the agent has identified, and a second position vector used to store the current global best solution among all agents; evaluating an objective function at the location of each agent, wherein the objective function represents a measure of registration quality between the set of transformed image windows and the reference image; and comparing the current global best solution found by all of the agents with an optimum solution, wherein if the global best solution is within a predetermined threshold of the optimum solution, then the global best solution represents the registrations; wherein the objective function is defined as: J = ( 1 N ⁢ ∑ k = 1 N ⁢ λ k ⁡ ( I max - d _ k ) , where N denotes the number of image windows, Σ represents a summation, λ k denotes the fraction of overlap of a kth image window with the reference image, I max is the maximum pixel value of the test image, and d k denotes the average absolute difference between the image window and the reference image.

10. A method for image registration utilizing particle swarm optimization as set forth in claim 9 , further comprising acts of applying a Gaussian kernel to the test image and the reference image to assist the convergence of the plurality of software agents.

11. A method for image registration utilizing particle swarm optimization as set forth in claim 10 , further comprising acts of applying a translation to each image window in the set of image windows, the set of image windows comprising a center-of-gravity, such that the center-of-gravity of the set of image windows coincides with the center of the reference image.

12. A method for image registration utilizing particle swarm optimization as set forth in claim 11 , further comprising acts of generating an image pyramid of both the test image and the reference image, each image pyramid having a plurality of levels comprising images, wherein each level of each image pyramid is an identical image having a different size and resolution.

13. A method for image registration utilizing particle swarm optimization as set forth in claim 12 , wherein the plurality of software agents are configured to explore each level of each image pyramid in search of the objective function, wherein the agents begin at a top level of each image pyramid and continue down each image pyramid until convergence is reached at a lowest level of each image pyramid.

14. A method for image registration utilizing particle swarm optimization as set forth in claim 13 , wherein the evaluation of the objective function is carried out at the same image pyramid levels using the reference image and a set of image windows extracted from the test image pyramid.

15. A computer program product for image registration utilizing particle swarm optimization, the computer program product comprising computer-readable instruction means stored on a non-transitory computer-readable medium that are executable by a computer having a processor for causing the processor to perform operations of: selecting a set of image windows from a test image; transforming each image window from the test image, such that a transformation of each image window aligns each image window with a reference image having a center, resulting in a set of transformed image windows; configuring a plurality of software agents to operate as a cooperative swarm to optimize an objective function, wherein each agent is assigned an initial velocity vector to explore a multi-dimensional solution space, where each agent is configured to perform at least one iteration, the iteration being a search in the multi-dimensional solution space for a potential objective function where each agent keeps track of a first position vector representing a current individual best solution that the agent has identified, and a second position vector used to store the current global best solution among all agents; evaluating an objective function at the location of each agent, wherein the objective function represents a measure of registration quality between the set of transformed image windows and the reference image; and comparing the current global best solution found by all of the agents with an optimum solution, wherein if the global best solution is within a predetermined threshold of the optimum solution, then the global best solution represents the registration; wherein the objective function is defined as: J = ( 1 N ⁢ ∑ k = 1 N ⁢ λ k ⁡ ( I max - d _ k ) , where N denotes the number of image windows, Σ represents a summation, λ k denotes the fraction of overlap of a kth image window with the reference image, I max is the maximum pixel value of the test image, and d k denotes the average absolute difference between the image window and the reference image.

17. A computer program product for particle swarm optimization as set forth in claim 16 , further operable for applying a Gaussian kernel to the test image and the reference image to assist the convergence of the plurality of software agents.

18. A computer program product for image registration utilizing particle swarm optimization as set forth in claim 17 , further operable for applying a translation to each image window in the set of image windows, the set of image windows comprising a center-of-gravity, such that the center-of-gravity of the set of image windows coincides with the center of the reference image.

19. A computer program product for image registration utilizing particle swarm optimization as set forth in claim 18 , further operable for generating an image pyramid of both the test image and the reference image, each image pyramid having a plurality of levels comprising images, wherein each level of each image pyramid is an identical image having a different size and resolution.

20. A computer program product for image registration utilizing particle swarm optimization as set forth in claim 19 , wherein the plurality of software agents are configured to explore each level of each image pyramid in search of the objective function, wherein the agents begin at a top level of each image pyramid and continue down each image pyramid until convergence is reached at a lowest level of each image pyramid.

21. A computer program product for image registration utilizing particle swarm optimization as set forth in claim 20 , wherein the evaluation of the objective function is carried out at the same image pyramid levels using the reference image and a set of image windows extracted from the test image pyramid.